Shim old easing parameters and provide a warning.

[gnarf]

jquery/jquery#2376 from @mr21 is going to be removing the "extra parameters" that have been passed to easing. We need to put a shim in place that can detect the arity of a easing function being >1 argument and provide a warning / shim the additional parameters back in.

[dmethvin]

Is this a common enough situation that it needs to be shimmed? Do we know of specific code that is depending on the extra parameters?

[gnarf]

the shim is like ~1 line of code? only needed in the "debug build" ?

[dmethvin]

Nothing in old jQuery versions uses the params? If it does I'd be concerned about noise that the user wouldn't understand.

[scottgonzalez]

Do we know of specific code that is depending on the extra parameters?

The easing plugin, which is still the #1 result on Google for "jquery easing".

[gnarf]

Nothing in old jQuery UI or jQuery uses it for a few versions back

[gnarf]

I also think it wont be very "noisy" - once per custom easing function that still uses the extra parameters

[mr21]

I already search where this parameters is used, and I didn't find anything, even after the first commit of jquery these parameters is useless :(

But I send another mail to the author of the jquery easing plugin, to upgrade it (I suggested my time to do it if he want).

[scottgonzalez]

Honestly, if a new version of that plugin is going to come out, it will be so different from the current version that it should just be deprecated in favor of jQuery UI's easings. We've already done the work of optimizing for size.

[mr21]

Yep, I saw your code, I like the way how it's written :)
But, maybe we can just do a copy/past of your version inside the plugin and that's it.
So, after that, the #1 result on google will not be a problem anymore, right?

[scottgonzalez]

While you have the legal right to do so (with the copyright and license terms staying in tact), there is really no reason to do that. How is that in any way better than pointing people to the original code?

Note that before the next jQuery UI release, easings will be in their own file.

[mr21]

there is really no reason to do that.

hmm, it's for the next users maybe, you said yourself when we search jquery easing on google, the first result is the plugin. So I thought the solution was to update the plugin directly.
I read in the jquery-ui file the original idea is from the plugin, so I believe in the fact you can do the same in return?

It's only a proposition, if you show me it's a bad move, I will not do it, but I not understand exactly why is not the good solution :(

[scottgonzalez]

hmm, it's for the next users maybe, you said yourself when we search jquery easing on google, the first result is the plugin. So I thought the solution was to update the plugin directly.

Or just ask the author to put a big notice on the page that says to use jQuery UI's easings.

I read in the jquery-ui file the original idea is from the plugin, so I believe in the fact you can do the same in return?

I'm not sure what you mean by this. We have an optimized version of all the easings. You'll be able to get them by themselves, with no dependencies other than jQuery. What is the benefit of pointing people to the old easing plugin? Where would that notice go? In what scenario is this beneficial for users?

It's only a proposition, if you show me it's a bad move, I will not do it, but I not understand exactly why is not the good solution :(

How is replacing 100% of the code in the easing plugin with a copy/paste from jQuery UI a good solution?

[mr21]

How is replacing 100% of the code in the easing plugin with a copy/paste from jQuery UI a good solution?

Only to fix the problem of the google's result. But I admit the notice on the page is another solution, yep, why not.

[dmethvin]

The problem we've historically tried to solve with Migrate is that people attempt to upgrade their jQuery core, but don't realize they need to upgrade several other dependencies or change their own code and things mysteriously break. Where possible, and it's usually possible, we both shim the now-broken functionality and give a warning about the fix that needs to be done. The warning should be clear and give the most likely causes.

So in this case it seems like the most commonly encountered situation by far will be someone using the old jQuery Easing plugin. If it's the first result on Google you can bet people are still using it and it would help them to get a warning. We'll need to punch .animate() to look at the function arity and wrap with a new function if it expects the old-style calling sequence, saving duration so we can reconstitute the old args at call time? Or am I over-thinking this?